INDEX

Explanations

words related to military ranks and terms

oai_token-act-pair · gpt-3.5-turbo

affirmative statements and expressions of fact or confirmation

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.7.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 whilst

-0.73

Âł

-0.72

 Whilst

-0.70

 civilisation

-0.67

".[

-0.67

isations

-0.66

↵Âł

-0.66

 Âł

-0.66

 msec

-0.65

''.

-0.65

POSITIVE LOGITS

 FedEx

0.66

Arizona

0.65

 nonprofits

0.65

adelphia

0.62

 Southwest

0.62

UPS

0.59

 Delaware

0.58

 Costco

0.58

Florida

0.57

"$

0.57

Activations Density 2.438%

words related to military ranks and terms

affirmative statements and expressions of fact or confirmation

No Comments

No Known Activations